Predicting Internalizing Problems in At-Risk Children and Adolescents

Tawnyea L. Bolme-Lake

DISSERTATION.COM

Boca Raton

Predicting Internalizing Problems in At-Risk Children and Adolescents

Copyright © 2007 Tawnyea L. Bolme-Lake All rights reserved.

Dissertation.com

Boca Raton, Florida USA • 2008

ISBN-10: 1-59942-659-5

ISBN-13: 978-1-59942-659-4

PREDICTING INTERNALIZING PROBLEMS IN

AT-RISK CHILDREN AND ADOLESCENTS

By

Tawnyea L. Bolme-Lake

A Dissertation Presented in Partial Fulfillment

of the Requirements for the Degree

Doctor of Philosophy

Capella University

August 2007

© Tawnyea L. Bolme-Lake, 2007

PREDICTING INTERNALIZING PROBLEMS IN

AT-RISK CHILDREN AND ADOLESCENTS

by

Tawnyea L. Bolme-Lake

has been approved

August 2007

WILLIAM CAMERON, Ph.D., Faculty Mentor and Chair

KELLEY CHAPPELL, Ph.D., Committee Member

ANTANAS LEVINSKAS, Ph.D., Committee Member

GARVEY HOUSE, Ph.D. Dean, School of Psychology

A Dissertation Presented in Partial Fulfillment

Of the Requirements for the Degree

Doctor of Philosophy

Capella University

August 2007

Abstract

Internalizing problems are common among adolescents. Poor outcomes such as

academic failure, substance misuse, and adult mental health problems have all been

linked to internalizing problems. Although the potential effects are serious, internalizing

disorders tend to be under-diagnosed and under-treated. To compound the problem,

research in the area of internalizing disorders continues to lag behind that of other

disorders. In the last ten years, however, research has indicated that relationships with

parents, gender, and self-esteem are factors associated with internalizing disorders. To

clarify the relationships between these factors, archival data was collected from an

electronic database in a school district in northeastern Minnesota. This database includes

the results of the Behavior Assessment System for Children-2; Self-Report of Personality

(BASC-2) of children and adolescents who have taken it as a part of a special education

evaluation. The BASC-2 is a norm-referenced questionnaire that measures emotions and

self-perceptions. Factorial analysis of variance was used determine whether the degree

of internalizing problems differ between relationships with parents, gender, and self-

esteem groups, reflected by scores on the Parent Relations, Self-Esteem, and Internalizing

Problems scales included in the BASC-2. Further, multiple regression procedures were

used to determine if the combination of the quality of relations with parents, gender, and

level of self-esteem predicts the degree of internalizing problems experienced by at-risk

children and adolescents. Contrary to past studies, results did not show gender significant

differences in the degree of internalizing problems reported. Results did, however,

indicate that the degree of reported internalizing problems was related to the quality of

parent relationships and self-esteem. Specifically, children and adolescents who reported

poor relationships with their parents reported a significantly greater degree of

internalizing problems than those who reported average or better relationships with their

parents. Likewise, children and adolescents who reported low self-esteem reported a

significantly greater degree of internalizing problems than those with average or better

self-esteem. In addition, gender, the quality of parent relations, and level of self-esteem

showed a predictive relationship with internalizing problems. The implications of these

findings, as well as directions for future research were discussed.

iv

Acknowledgments

There are many individuals to who I wish to acknowledge for making this project

possible. Faculty from Capella University, professionals from the public school district

from which the data was gathered, as well as my friends and family all played an

important role in the success of this project. Without their support, completion would not

have been possible.

First, I would like to extend my sincere gratitude to the members of my

dissertation committee. Dr. William Cameron, mentor and dissertation chair, provided

support, guidance, and expertise. Dr. Antanas Levinskas and Dr. Kelley Chappell also

provided a good deal of support and encouragement throughout this process.

Second, I would like to thank several top-notch professionals from the public

school district from which the data was gathered. Specifically, Dr. Keith Dixon and

Marci Hoff granted me permission to use data from their school district, and Angela Sepp

and Cody Chamberlain assisted in the electronic data collection.

Third, I would like to thank the "Kids Table". Their support made my educational

goals seem attainable; their intellect challenged me and made me proud to be a part of

such an accomplished group.

Fourth, I would like to thank my nearest and dearest friends for sticking with me

throughout this journey. My "sisters", Margie and Lisa, my "couple friends", Dave and

Shelley, Will and Sue, Chuck and Vicky… all kept me laughing, smiling, and having fun

throughout the process.

And finally, I would like to give a heartfelt thanks to my family. My husband Ron

has provided unconditional love and unwavering support throughout all of my

v

educational endeavors. His belief in me gave me the strength to see this to completion.

My amazing children Kaitlyn and Alek provided daily inspiration. When difficulties were

encountered, their smiling faces and sincere acts of love kept everything in perspective.

Brother D and sister-in-law MP, Baby Bolme, Bob, Char, Rob, Karie and the kids… each

provided helpful support in one way or another. My parents, Rich and Donna provided

me with the on-going belief that anything is possible. The often-repeated phrase "you can

do anything you put your mind to" resonates in me and has guided me throughout my

life. Thank you!

vi

Table of Contents

Acknowledgments iv

List of Tables viii

List of Figures ix

CHAPTER 1. INTRODUCTION 1

Introduction to the Problem 1

Background of the Study 4

Statement of the Problem 4

Purpose of the Study 5

Significance of the Study 6

Nature of the Study 6

Research Questions 6

Definition of Terms 7

Assumptions and Limitations 10

Organization of the Remainder of the Study 12

CHAPTER 2. LITERATURE REVIEW 14

Introduction to the Literature Review 14

Critical Review of the Relevant Literature 14

Rationale for Methodology, Design, and Measurement 22

Chapter Summary 30

CHAPTER 3. METHODOLOGY 32

Restatement of Purpose 32

Research Design 32

vii

Target Population 32

Selection of Documents 33

Definition of Variables 33

Instruments 35

Data Collection and Procedures 36

Research Questions and Hypotheses 37

Data Analysis 39

Expected Findings 40

CHAPTER 4. RESULTS 42

Organization of the Chapter 42

Characteristics of the Sample 42

Tests of Hypotheses 1 – 3 44

Test of Hypothesis 4 56

Summary of Results 64

CHAPTER 5. DISCUSSION 66

Organization of the Chapter 66

Overview of the Results 66

Discussion and Interpretation of Findings 67

Strengths 73

Limitations 74

Recommendations for Future Research 76

Summary and Conclusion 78

REFERENCES 79

viii

List of Tables

Table 1. Descriptive Statistics of Each Group 43

Table 2. Descriptive Statistics for Internalizing Problems at Each Variable Combination 44

Table 3. Skewness and Kurtosis of Internalizing Problems 46

Table 4. Tests of Normality on the Dependent Variable Internalizing Problems 47

Table 5. Levene's Test of Equality of Error Variances 48

Table 6. Analysis of Variance for Internalizing Problems 54

Table 7. Correlations Between Variables 57

Table 8. Summary of Variables Not Yet Entered Into Regression Model 58

Table 9. Summary of Stepwise Regression Analysis for Variables Predicting 62 Internalizing Problems Table 10. Analysis of Variance, Change in Internalizing Problems 63

ix

List of Figures

Figure 1. Histogram of Mean Internalizing Problems Scores 46

Figure 2. Box Plot of Mean Internalizing Problems Scores by Parent Relations Group 49

Figure 3. Box Plot of Internalizing Problems Scores by Gender Group 50

Figure 4. Box Plot of Internalizing Problems Scores by Self-Esteem Group 51

Figure 5. Results of Analysis of Variance 55

Figure 6. Scatter Plot of Standardized Residuals by Standardized Predicted Values 59

Figure 7. Histogram of Standardized Residuals 60

Figure 8. Plot of Residual Values 61

Figure 9. Visual Depiction of Multiple Regression Results 64

1

CHAPTER 1. INTRODUCTION

Introduction to the Problem

Mental health conditions can be divided into two broad categories, internalizing

disorders and externalizing disorders. This approach to classification is based on the

empirical work of Achenbach and colleagues (Achenbach, 1966, 1985; Achenbach &

Edelbrock, 1978; Achenbach & McConaughy, 1996). Using multivariate factor analysis,

two large groups of conditions were identified. Inner-directed, over-controlled behaviors

that cause emotional distress in the self were classified as internalizing disorders. In

contrast, behavior disorders that create conflicts within the environment or with others

were categorized as externalizing disorders (Reynolds, 1990).

Internalizing disorders such as anxiety and depression are common among

children and adolescents. In fact, the estimated prevalence of anxiety disorders in

children and adolescents is 13% (U.S. Surgeon General, 1999). In addition, some studies

suggest that as many as 20% of adolescents will experience at least one episode of

clinically significant depression in their lifetimes (Birmaher et al., 1996; Garber, 2000;

Lewinsohn & Essau, 2002). Both of these disorders are included within the cluster of

internalizing disorders (Achenbach, 1966, 1985; Achenbach & Edelbrock, 1978;

Achenbach & McConaughy, 1996).

Although researchers cite high prevalence rates for these disorders, some argue

that statistics actually underestimate their true incidence (Reynolds, 1990). This occurs

because these disorders present symptoms that are not always observable (Laurent &

2

Landau, 1993). In addition, the inner-directed nature of these disorders does not affect

others as the outer-directed externalizing disorders tend to, nor do they present behavior

management challenges for parents, teachers, or mental health professionals. As a result,

internalizing disorders are often overlooked (Reynolds, 1992).

Even when internalizing disorders are diagnosed, research suggests that they are

often under-treated. Wu et al. (1999) found that youngsters with externalizing problems

are likely to receive services through mental health organizations and schools. In contrast,

youngsters with internalizing disorders, such as depression, are more likely to receive

services solely in their schools. In other words, children and adolescents with

internalizing-type disorders may get less treatment than those with other types of

disorders. This discrepancy occurs because externalizing disorders tend to cause more

distress in others, causing parents to seek additional support from community agencies

(1999).

Untreated internalizing disorders are related to serious problems. Academic

failure is one of the problems associated with internalizing disorders (National

Association of School Psychologists, 2002; Rapport, Denney, Chung, & Hustace, 2001;

Reynolds, 1992). Rapport et al. (2001) conducted a study using a sample of 325 children

and adolescents ages 5 - 7. Measures such as intelligence, classroom performance,

internalizing behavior, short-term memory, and vigilance were used to examine the

relationship between internalizing problems and classroom performance. Results

indicated that internalizing behavior, defined in this study as anxiety, depression, and

withdrawal, contributed significantly to the prediction of classroom performance "over

and above the effects of intelligence" and all other factors (p. 548).

3

Substance misuse is another problem related to internalizing disorders (Kubik,

Lytle, Birnbaum, Murray, & Perry, 2003; Lillehoj, Trudeau, Spoth, & Wickrama, 2004;

Loeber, Stouthamer-Loeber, & White, 1999; Wang, Fitzhugh, & Westerfield, 1994).

Kubik et al. (2003) surveyed 3,621 12 and 13 year-olds in 16 different middle schools in

Minnesota. Information was collected for gender, age, race/ethnicity, depressive

symptoms, smoking, alcohol use, and use of marijuana and inhalants. Results showed that

in both boys and girls, depressive symptoms were strongly associated with monthly

alcohol and inhalant use. Monthly smoking and heavy drinking were associated with

depressive symptoms in girls only. This suggests that the association between substance

use and internalizing disorders is a concern in young adolescents as well as older

adolescents. It also underscores the existence of gender differences.

Internalizing disorders in childhood and adolescence are also associated with

mental health problems later in adulthood (Pine, Cohen, Cohen, & Brook, 1999; National

Association of School Psychologists, 2003). Pine et al. analyzed a sample of 776

adolescents with depressive symptoms who had psychiatric evaluations completed in

1983, 1985, and 1992. Results showed that adolescent depressive symptoms strongly

predicted adult major depression. In fact, adolescents with clinical depression were 2 - 3

times more likely to have at least one major depressive episode as an adult.

Despite the high prevalence of these disorders and the associated detrimental

effects, research in the area of internalizing disorders lags behind in comparison to

research in the area of externalizing disorders (Compton, Burns, Egger, & Robertson,

2002). This lag is especially evident with respect to the child and adolescent population.

4

Background of the Study

The 1970s marked the beginning of research on internalizing disorders. During

this time, research efforts focused on the nature and treatment of these disorders in adults

(Reynolds, 1992). However, it wasn't until the 1980s that the study of internalizing

disorders trickled down to the child and adolescent population. Prior to that, children and

adolescents with internalizing symptoms were viewed as going through normal, but

difficult, developmental stages in their lives. The common belief was they would simply

"grow out" of their symptoms (1992). Reynolds argued that this growing attention to

child and adolescent internalizing disorders stemmed in part from a rapidly increasing

suicide rate among adolescents in the 1950s, 1960s, and 1970s. The publication of the

Diagnostic and Statistical Manual of Mental Disorders-III in 1980 also sparked interest in

the phenomenon of internalizing disorders of childhood and adolescence (1992).

Statement of the Problem

Since the 1980s, studies have identified that certain factors, such as gender, put

some individuals at a higher risk of developing internalizing disorders. Specifically,

many studies indicate that girls are far more likely to develop internalizing disorders than

boys (Leadbeater, Blatt, & Quinlan, 1995; Crawford, Cohen, Midlarsky, & Brook, 2001;

Kubik, et al., 2003; Jose & Ratcliff, 2004; Ronnlund & Karlsson, 2006). Researchers

have also identified certain factors that may make some children and adolescents less

vulnerable to internalizing disorders. Quality relationships between adolescents and their

parents, as well as high self-esteem have been implicated as protective factors against

problems in psychological adjustment (Schweitzer, Seth-Smith, & Callan, 1992; Delaney,

5

1996; Bryne, 2000; Kliewer, Murrelle, & Meja, 2001; Erkolahti, Ilonen, Saarijavi, &

Terho, 2003; Marsh, Parada, & Ayotte, 2004; Reid, 2004; Manders, Scholte, Janssens, &

De Bruyn, 2006; Margolin, 2006; Ronnlund & Karlsson, 2006).

Although certain risk factors and protective factors have been identified, research

has been limited to simple relationships between these variables; no research to date has

examined the possible interactions between these factors. In addition, researchers have

not explored the potential predictive relationship between this combination of factors and

the degree of internalizing problems reported by children and adolescents. This study fills

those gaps in the literature.

Purpose of the Study

Given the high prevalence of internalizing disorders, the associated detrimental

effects, and the relative lack of research on internalizing disorders in comparison to

externalizing disorders, additional research is needed to identify possible factors or

combination of factors that put adolescents at a higher risk for developing these disorders.

This study examines whether girls experience greater internalizing problems in

comparison to boys, and whether the lack of quality relationships with parents and low

self-esteem are associated with a high degree of internalizing problems. It will also

examine interaction effects. Additionally, it employs multiple regression procedures to

determine if any of these factors or combination of factors, have a predictive relationship

with internalizing problems.

6

Significance of the Study

This study identifies the potentially complex interplay between the factors of

quality of relations with parents, gender, level of self-esteem, and degree of internalizing

problems. It is expected that these findings will be useful in identifying unique

interactions between these variables. As a result, a profile of characteristics that put

children and adolescents at a higher risk for development of internalizing disorders will

be identified. Identification of risk factors may, in turn, help guide interventions for these

groups.

Nature of the Study

This study uses a causal-comparative factorial design to examine how the quality

of parent relationships, gender, and level of self-esteem are related to internalizing

problems in the at-risk population of children and adolescents. Specifically, this study

uses archival data to determine whether there are interactions between and among these

factors. In addition, this research uses multiple regression procedures to examine whether

any of these variables or combination of variables can predict future internalizing

problems in at-risk children and adolescents.

Research Questions

1. Is there a statistically significant difference in the degree of internalizing

problems between the two Relations with Parents groups (Average/Above

Average, At-Risk/Clinically Significant) as measured by the BASC-2?

7

2. Is there a statistically significant difference in the degree of internalizing

problems between the two gender groups (Male, Female) as measured by the

BASC-2?

3. Is there a statistically significant difference in the degree of internalizing

problems between the two Self-Esteem groups (Average/Above Average, At-

Risk/Clinically Significant) as measured by the BASC-2?

4. Does the combination of quality of relations with parents, gender, and level of

self-esteem have a predictive relationship with the degree of internalizing

problems reported by at-risk children and adolescents?

Definition of Terms

There are several terms related to this study that require further explanation.

These terms will be used throughout the remainder of the study as defined in this section.

At-Risk Children and Adolescents

For the purpose of this research, at-risk children and adolescents will refer to

youngsters between the ages of 8-18 that either receive special education services as a

result of an identified emotional/behavioral disorder or have been referred for a special

education evaluation because an emotional/behavioral disorder is suspected. This

population is the focus of this study.

At-Risk/Clinically Significant Relations with Parents

According to Reynolds and Kamphaus (2004), T-scores of 31 - 40 on any

adaptive measure, such as Parent Relations, are considered At-Risk. Scores that fall

within the At-Risk range indicate "the presence of significant problems" or "may signify

8

potential or developing problems that need to be monitored carefully" (p. 16). Likewise,

T-scores of 30 and below are considered Clinically Significant in the area of Relations

with Parents and "denote a high level of maladaptive behavior" (p. 16). The At-

Risk/Clinically Significant Parent Relations group will consist of children and

adolescents who reported problems within the parent-child relationship.

At-Risk/Clinically Significant Self-Esteem

In the area of Self-Esteem, T-scores of 31 - 40 are considered At-Risk on the

BASC-2 (Reynolds & Kamphaus, 2004). According to Reynolds and Kamphaus, any

adaptive measure, including Self-Esteem, that reflects scores that fall within the At-Risk

range indicate "the presence of significant problems" or "may signify potential or

developing problems that need to be monitored carefully" (p. 16). Likewise, T-scores 30

and below are considered Clinically Significant. Scores that fall within the Clinically

Significant range "denote a high level of maladaptive behavior" (Reynolds & Kamphaus,

2004, p. 16). Therefore, children and adolescents in the At-Risk/Clinically Significant

Self-Esteem group represent individuals with low self-esteem.

Average/Above Average Relations with Parents

In the area of Relations with Parents, T-scores of 60 and above are considered

High or Very High and indicate very positive relations between the parents and the child

or adolescent. T-scores of 41 - 59 are considered Average and reflect average quality

relationships between the child or adolescent and his or her parents (Reynolds &

Kamphaus, 2004). Thus, the Average/Above Average Parent Relations group consists of

youngsters who reported average or better relationships with their parents.

9

Average/Above Average Self-Esteem

In the area of Self-Esteem, T-scores of 60 and above are considered High and

Very High on the BASC-2 and reflect a level of self-esteem that is better than average.

Likewise, T-scores of 41 - 59 are considered Average, and reflect an average level of

self-esteem (Reynolds & Kamphaus, 2004). Thus, children and adolescents with

Average/Above Average Self-Esteem reflect individuals with average or better self-

esteem.

Behavior Assessment System for Children; Second Edition (BASC-2)

The BASC-2 is a multi-method, multidimensional assessment tool that evaluates

behavior and self-perceptions of children and adolescents (Reynolds & Kamphaus, 2004).

This is one of the evaluation tools used in the school district under study when

considering the special education eligibility. It is also the tool that was used to measure

the quality of relationships with parents, level of self-esteem, as well as the degree of

internalizing problems in this study.

Emotional/Behavioral Disorder (EBD)

According to National Association of School Psychologists (2002), EBD refers to

"a condition in which behavioral or emotional responses of an individual in school are so

different from his/her generally accepted, age-appropriate, ethnic or cultural norms that

they adversely affect performance in such areas as self care, social relationships, personal

adjustment, academic progress, classroom behavior, or work adjustment" (para. 3). All of

the students whose BASC-2 results were used for this study have been evaluated and

identified as having an EBD, another disability, or were suspected of having an EBD.

10

Internalizing Problems

Internalizing disorders are a group of disorders that are described as inner-

directed and over-controlled (Reynolds, 1992). The Internalizing Problems composite

score on the BASC-2 Self-Report of Personality includes the scales of Atypicality

(tendency to behave in a manner considered odd or strange), Locus of Control, Social

Stress, Anxiety, Depression, and Sense of Inadequacy. Reynolds and Kamphaus (2004)

consider the Internalizing Problems composite on the BASC-2 as a "broad index of

inwardly directed distress."

Parent Relationships

In this study, the quality of the relationship between the child or adolescent and

the parents was measured using the Relations with Parents scale on the BASC-2.

According to Reynolds and Kamphaus (2004), this scale "surveys the individual's

perception of being important in the family, the status of the child-parent relationship,

and the child's perception of the degree of parental trust and concern" (p. 78).

Self-Esteem

The level of self-esteem was measured in this study using the Self-Esteem scale

on the BASC-2. This scale evaluates the adolescent's satisfaction with one's self,

physically and globally (Reynolds & Kamphaus, 2004).

Assumptions and Limitations

There are several key assumptions embedded within this study that are worth

noting. The first group of assumptions is related to the quantitative philosophy of this

study. With regard to ontology, this research assumes a single reality rather than multiple

11

realities. The epistemology assumes that the researcher and the data are independent

entities. The axiology of this study assumes that the researcher's values will not impact

the final results. It is also assumed that at least some degree of generalizations will be

possible as a result of the study and that causes and effects exist in a linear manner.

Finally, this study assumes a deductive method of logic.

The second group of assumptions is related to the use of the BASC-2 as the tool

of measurement used in this study. First, it is assumed that the school professionals that

administered the BASC-2 as a part of the special education evaluation process followed

proper testing and scoring procedures. Second, it is assumed that the students that

completed the BASC-2 answered the questions openly and honestly. Third, it is assumed

that the demographics of the students that took that BASC-2 in the school district under

study roughly match the demographics of the students included in the BASC-2 norm

group.

The limitations of this study should also be taken into account when making

generalizations from the final results. The first group of limitations is related to the

concept of internalizing disorders. Although the distinction between internalizing

disorders and externalizing disorders has empirical support (Achenbach, 1966; 1985), it

is important to note that not every mental health condition falls neatly into these two

categories. Rather, internalizing disorders often have symptoms that overlap with other

externalizing disorders (McConaughy & Skiba, 1993). Second, internalizing disorders

and externalizing disorders are often co-morbid (1993). Finally, not all researchers agree

which specific mental health conditions fall within the category or internalizing disorders

(Reynolds, 1992). All of these factors affects the interpretation of the final results.